LightSpin Technologies, Inc., proposes a Phase II SBIR to develop a new photodetector component (the GaAs Photomultiplier Chip) for ultrafast TOF-PET applications. The goal is to build a solid-state photodetector with a single photon timing resolution of 50 ps rms in Phase II and 25 ps rms in a follow-on effort. A single photon timing jitter of 25 ps rms would be about 4 better than the current state-of-the-art (photomultiplier tubes and silicon photomultiplier devices for TOF-PET). Furthermore, the GaAs Photomultiplier Chip will achieve 25% quantum efficiency in Phase II and 50% in a follow-on effort, further improving the timing resolution in TOF-PET (compared to the 10 - 20% quantum efficiency of today's solutions). In Phase I, we demonstrated the ability to achieve single-photon rise times as fast as 100 picoseconds. In Phase II, these results will be extended to improved timing jitter, higher single photon detection efficiencies, and larger active areas. The breakthrough stems from LightSpin's innovation enabling compound semiconductors to be used for single photon detectors, combined with optimization of the monolithically integrated passive quench circuitry. GaAs is inherently faster than silicon due to its 5 higher mobility (reducing transit time spread) and it sharp avalanche breakdown characteristics (reducing jitter in the avalanche gain process). The end result of the Phase II effort will be an 8 8 array of 16 mm2 Photomultiplier Chips, coupled to a matching array of scintillators provided by Saint-Gobain Crystals and including high performance electrical readout circuitry and digitization electronics. The performance for ultrafast TOF-PET applications will be evaluated by Prof. Joel Karp at the University of Pennsylvania, and benchmarked against state-of-the-art alternatives. The Phase II goal is to demonstrate a TOF-PET coincidence timing resolution of 200 psec, which is sufficiently fast to enable a substantial improvement in PET performance in brain and breast imaging applications. Additionally, the Phase II project is expected to pave the way for a follow-on effort to reduce the coincidence timing resolution to below 50 psec, providing an order of magnitude improvement compared to currently available commercial TOF-PET systems.